Mechanism of MTO-catalyzed deoxydehydration of diols to alkenes using sacrificial alcohols

Article abstract of DOI:10.1021/om400127z

Catalytic deoxydehydration (DODH) of vicinal diols is carried out employing methyltrioxorhenium (MTO) as the catalyst and a sacrificial alcohol as the reducing agent. The reaction kinetics feature an induction period when MTO is added last and show zero-order in [diol] and half-order dependence on [catalyst]. The rate-determining step involves reaction with alcohol, as evidenced by a KIE of 1.4 and a large negative entropy of activation (ΔS^? = -154 ± 33 J mol^-1 K ^-1). The active form of the catalyst is methyldioxorhenium(V) (MDO), which is formed by reduction of MTO by alcohol or via a novel C-C bond cleavage of an MTO-diolate complex. The majority of the MDO-diolate complex is present in dinuclear form, giving rise to the [Re]^1/2 dependence. The MDO-diolate complex undergoes further reduction by alcohol in the rate-determining step to give rise to a putative rhenium(III) diolate. The latter is the active species in DODH extruding stereoselectively trans-stilbene from (R,R)-(+)-hydrobenzoin to regenerate MDO and complete the catalytic cycle.

Full text of DOI:10.1021/om400127z

Article
pubs.acs.org/Organometallics
Mechanism of MTO-Catalyzed Deoxydehydration of Diols to Alkenes
Using Sacrificial Alcohols
†
†,§
‡
†
†
†
Shuo Liu, Aysegul Senocak, Jessica L. Smeltz, Linan Yang, Benjamin Wegenhart, Jing Yi,
†
‡
,†
̈
Hilkka I. Kenttamaa, Elon A. Ison, and Mahdi M. Abu-Omar*
†
Brown Laboratory and Department of Chemistry, Purdue University, 560 Oval Drive, West Lafayette, Indiana 47907, United States
‡Department of Chemistry, North Carolina State University, Raleigh, North Carolina 27695, United States
S Supporting Information
*
ABSTRACT: Catalytic deoxydehydration (DODH) of vicinal
diols is carried out employing methyltrioxorhenium (MTO) as
the catalyst and a sacrificial alcohol as the reducing agent. The
reaction kinetics feature an induction period when MTO is
added last and show zero-order in [diol] and half-order
dependence on [catalyst]. The rate-determining step involves
reaction with alcohol, as evidenced by a KIE of 1.4 and a large
‡
−1 −1
negative entropy of activation (ΔS = −154 ± 33 J mol K ).
The active form of the catalyst is methyldioxorhenium(V)
(
MDO), which is formed by reduction of MTO by alcohol or
via a novel C−C bond cleavage of an MTO-diolate complex.
The majority of the MDO-diolate complex is present in
dinuclear form, giving rise to the [Re] dependence. The
1
/2
MDO-diolate complex undergoes further reduction by alcohol in the rate-determining step to give rise to a putative rhenium(III)
diolate. The latter is the active species in DODH extruding stereoselectively trans-stilbene from (R,R)-(+)-hydrobenzoin to
regenerate MDO and complete the catalytic cycle.
he development of renewable energy sources has become
an important research field because of dwindling fossil
hydrogenation of biomass-derived glycerol, polyols, and sugar
9,10
T
alcohols.
fuel supplies and concerns over greenhouse gas emissions.
Biomass represents an attractive and sustainable platform for
the production of liquid fuels and valuable chemicals through1
deoxygenation of biomass-derived polyols and sugar alcohols.
A promising route toward this goal is catalytic deoxydehydra-
tion (DODH) of vicinal diols to alkenes employing rhenium-
based catalysts. Nonetheless, there are only a few studies
dealing with DODH in comparison with the reverse reaction,
dihydroxylation of alkenes, which is a widely used reaction in
In addition, Schlaf and Bullock have pioneered the use of
organometallic ruthenium catalysts for the deoxygenation of
1
1
alcohols. Srivastava’s group has also utilized (Cp*Ru(CO)
)
2 2
for hydrodeoxygenation (HDO) and hydrocracking of diols
and epoxides.
12
Despite the recent flurry of studies dealing with catalytic
deoxydehydration of diols, there have been limited kinetic
investigations aimed at understanding the reaction mechanism
under relevant catalytic conditions. Previous studies by our
group and others had indicated two possible reaction pathways.
2
organic synthesis and industry. Two organometallic oxorhe-
5
,7,9,13
These are illustrated in Scheme 1.
Pathway A invokes
nium complexes have been reported to catalyze the
deoxygenation of epoxides and deoxydehydration of diols to
formation of rhenium(VII) diolate, which undergoes reduction
3
V
to form Re diolate, followed by alkene extrusion to complete
alkenes. Cook and Andrews reported catalytic deoxydehydra-
V
the catalytic cycle. Alternatively, pathway B proceeds by MTO
reduction first with the sacrificial alcohol to afford
methyldioxorhenium(V) (MDO), which coordinates the diol
tion of diols with [Cp*Re (O)(diolate)] employing phos-
4
phines as reductants/oxygen atom acceptors. Our group
studied the conversion of diols and epoxides to alkenes using
V
to give Re diolate. Both cycles share a common alkene-
molecular hydrogen (H ) and methyltrioxorhenium(VII)
2
V
5
forming step from Re diolate (MDO·diolate).
(
MTO) as the catalyst. Subsequently, Bergman and co-
From systematic experimentation, we have come to the
realization that employing sacrificial alcohols as reductants
represents a good system for detailed mechanistic and kinetic
study of the MTO-catalyzed DODH reaction (eq 1). The
workers employed sacrificial alcohols as the hydrogen source
6
with Re (CO) as a catalyst, and the Nicholas group used
2
10
sulfite as a reducing equivalent in the presence of several
7
oxorhenium catalysts. Fernandes et al. reported on rhenium-
catalyzed deoxygenation of epoxides without the use of a
8
reducing agent. Most recently and independently, our group as
Received: February 13, 2013
well as Toste’s reported rhenium-catalyzed DODH and transfer
©
XXXX American Chemical Society
A
dx.doi.org/10.1021/om400127z | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Scheme 1. Proposed Literature Mechanisms for DODH of
Vicinal Diols to Alkenes Catalyzed by MTO
complexes (4 and 2 in Scheme 1) are observed, we present
compelling evidence based on careful kinetic analysis that the
active form of the catalyst that affords alkene is most likely
rhenium(III) diolate. The catalytic reaction gives a primary
kinetic isotope effect on the secondary alcohol 3-octanol (kH/
kD = 1.4), which is consistent with the reduction step of
rhenium(V) diolate 4 as part of the rate-determining step(s).
Our mechanistic insight should aid henceforth in the design of
appropriate rhenium and other metal catalysts for DODH of
diols and polyols.
RESULTS AND DISCUSSION
■
Reactions of Glycols with Sacrificial Alcohols. We
examined the reactivity of several glycols and epoxides with
MTO utilizing 3-octanol as both the solvent and reductant. The
results are listed in Table 1. A mixture of a given glycol (1
equiv) with excess 3-octanol (10 equiv) in the presence of
MTO (2 mol %) as the catalyst at 140 or 170 °C gave 100%
conversion and good to moderate yields of the corresponding
alkene along with 3-octanone, with reaction times ranging from
2
4 to 60 min. Aromatic diols were found to be more reactive
reaction proceeds smoothly under ambient pressure, open to
air, in comparison to other systems that require high-pressure
reactors. Here we present a thorough study of the rhenium-
catalyzed DODH mechanism focusing on (R,R)-(+)-hydro-
benzoin as the representative glycol substrate. We have chosen
this substrate to study due to the stability of its product, trans-
stilbene, under the reaction conditions. The reaction is zero-
order in terms of (R,R)-(+)-hydrobenzoin and has an induction
period when MTO is employed as the starting catalyst. This
induction period is significantly reduced when MDO is used as
the starting catalyst. These results indicate that MDO is indeed
part of the catalytic cycle, while MTO is not. Furthermore, the
than aliphatic diols. In addition, trans-1,2-cyclooctanediol did
not react, while cis-1,2-cyclooctanediol converted smoothly to
the alkene; this stereoselectivity has been noted previously for
5
−7
the H
-driven deoxygenation and other DODH.
Other
2
secondary alcohols such as nonanol and cyclooctanol can also
be used for DODH of vicinal diols and afford good alkene
yields (Table 1, entries 10−12). The slower reaction time with
5-nonanol can be attributed to steric hindrance. Interestingly,
primary and benzylic alcohols such as 1-heptanol, benzyl
alcohol, 1-phenylethanol, and diphenylmethanol are not
effective as sacrificial alcohols in this reaction system. 1-
Heptanol did not afford 1-heptanal, but it can be used as a
solvent for the conversion of (R,R)-(+)-hydrobenzoin to trans-
stilbene per eq 2. In the absence of a suitable sacrificial alcohol,
kinetic dependence on the catalyst concentration [Re] is half-
T
order. Even though both rhenium(VII) and rhenium(V) diolate
a
Table 1. MTO-Catalyzed DODH of Vicinal Diols and Deoxygenation of Epoxides Using Sacrificial Secondary Alcohols
a
b
Reactions were carried out at 2.5 mmol of glycol with 10 equiv of sacrificial alcohol and 2 mol % MTO. Refers to the time when the yield of the
c
major product reached maximum. All conversions are 100%. Yields of the major product(s) were determined by NMR using diphenylmethane as an
internal standard. Please refer to Supporting Information and ref 9 for details.
d
B
dx.doi.org/10.1021/om400127z | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
(
method A; see Experimental Methods section for details).
Aliquots were drawn at different times, quenched over ice, and
analyzed by NMR and/or GC for determination of hydro-
benzoin conversion and product yields. Figure 1a shows a
representative kinetic profile for the disappearance of hydro-
benzoin. The kinetics are characterized by an initial induction
period (ca. 200 s) and zeroth-order dependence on [hydro-
benzoin], as evidenced by the constant rate to more than 90%
conversion. The zero-order dependence on substrate (R,R-
hydrobenzoin) is indicative of a rate-determining step that
involves active-catalyst formation or alkene extrusion.
Another method for examining the kinetics of the reaction is
to add the substrate, hydrobenzoin, last to a heated solution of
MTO (catalyst) and 3-octanol (method B). Here, a mixture of
3-octanol (10 equiv relative to diol) with a catalytic amount
MTO was heated to 140 °C for 10 min, at which point the
the DODH reaction proceeds using the substrate hydrobenzoin
as a reductant. The product distribution includes trans-stilbene
and benzaldehyde as major products and benzil and 1,2-
diphenylethanone as minor products (eq 2; see Supporting
6
,12
Information).
Benzyl alcohol undergoes etherification to
give dibenzyl ether, and 1-phenylethanol is dehydrated by
13a,14
MTO to give styrene.
Kinetic Measurements. Kinetic measurements on the
progress of reaction were collected by heating a mixture of
hydrobenzoin with excess 3-octanol (10 equiv) to 140 °C. The
reaction was initiated by addition of 2, 5, or 10 mol % of MTO
Figure 1. (a) Kinetic profile for DODH reaction of hydrobenzoin in 3-octanol at 140 °C. Conditions: 2.5 mmol of hydrobenzoin with ∼25 mmol of
-octanol preheated to 140 °C, MTO (0.05 mmol) added last (method A); (b) kinetic profile for DODH of hydrobenzoin. Same conditions as in
a) except hydrobenzoin was added last after incubation of MTO in 3-octanol at 140 °C for 10 min; (c) kinetic profile for three consecutive catalytic
reactions. First in circles represents 2.5 mmol of hydrobenzoin with 50 mmol of 3-octanol preheated to 140 °C with MTO (0.05 mmol) added last
induction period omitted). Second and third in squares and hatched-squares, respectively, were initiated by addition of 2.5 mmol of hydrobenzoin
ca. 10−15 min after complete conversion of the substrate from the previous addition. Subsequent additions of substrate are designated by the arrows.
3
(
(
C
dx.doi.org/10.1021/om400127z | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
1
Figure 2. (a) H NMR spectrum of MDO·(PPh ) in CDCl . The peak at ∼2.8 ppm refers to the methyl group on Re (Re−CH ); the y axis is in
3
2
3
3
arbitrary units. (b) Kinetic profile when using 2 mol % MDO·(PPh ) as the catalyst (preheated hydrobenzoin with 10 equiv of 3-octanol at 140 °C,
catalyst added last, method A). (c) H NMR spectrum of MDO·(PCy ) in CDCl . The peak at ∼2.9 ppm refers to the methyl group on Re (Re−
CH ). (d) Kinetic profile when using 10 mol % MDO·(PCy ) as the catalyst, using method A.
3
2
1
3
2
3
3
3 2
solution turned dark in color. The appropriate amount of
hydrobenzoin was then added, and the reaction progress
followed by collecting aliquots and analyzing them by NMR
and GC. Figure 1b displays a typical kinetic profile for method
B. The induction period was almost absent, and the observed
rate (slope of the graph) was approximately 1.5× that in Figure
first cycle (omitted here), but no induction period was seen for
the second and third cycles. Additionally, all three consecutive
runs displayed zero-order dependence on [hydrobenzoin]. The
rates of the second and third catalytic cycles were ca. 1.5× that
of the first cycle, which is the rate with the active catalyst. In the
same way, the rate collected by method B with hydrobenzoin
added last (10 equiv of 3-octanol) is also 1.5 times faster than
the rate collected by method A (10 equiv also) with MTO
added last. All of these results are consistent with our notion
that the active catalyst is generated by the reaction of MTO
with 3-octanol, and there is little loss of activity even after 150
turnovers. Since the reactant concentrations of methods A and
B are not the same, the rates (slopes) cannot be compared
directly.
1a, where the catalyst was added last. It should be noted that
the order of addition had no effect on the zeroth-order
dependence on [hydrobenzoin]. Both profiles in Figure 1a and
b show a constant rate throughout the reaction. The lack of
induction period in Figure 1b is consistent with formation of
the active catalyst from a reaction between MTO and 3-octanol.
However, neither mechanism in Scheme 1 explains the
observation of an induction period.
To probe the effect of reaction time on catalyst activity, we
collected kinetic data for multiple reaction cycles starting with
0 equivalents of 3-octanol to hydrobenzoin. Since there would
Extended reaction times of MTO with 3-octanol at 140 °C
result in slow formation of a black precipitate, which has been
identified as Re nanoparticles (see the Supporting Information
2
15
be more samples taken from the reaction mixtures, the amount
of 3-octanol was increased to 20 equivalents. Ten to fifteen
minutes after completion of the first reaction, another
equivalent of hydrobenzoin was added to run a second catalytic
cycle. After completion of the second cycle, a third catalytic
cycle was commenced in the same manner. The results are
shown in Figure 1c. An induction period was observed for the
for details). The formed Re NPs can be isolated by filtration,
and their activity was compared with that of the supernatant.
The Re NPs were found to be much less active than the
supernatant (filtrate). These findings are consistent with the
active catalyst being homogeneous and not heterogeneous.
Further validation of this premise was obtained from the Hg(0)
poison test. Liquid mercury poisons the surface of heteroge-
D
dx.doi.org/10.1021/om400127z | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
1
6
neous metal catalysts. If the reaction is not affected by the
addition of mercury, such a result suggests a homogeneous
reaction and vice versa. The kinetics of the reaction was
compared in the presence of 10 equivalents of mercury present
at the beginning of the reaction prior to catalyst addition. The
rate of reaction and product selectivity were unaltered by the
presence of mercury, which indicates that the active catalyst is
homogeneous.
Experimental Methods section and Figure 8). The unique ion
cluster pattern based on Cl and Re isotope distributions
confirms the composition of the ion. A small amount of
deprotonated Re(VII) diolate, probably formed via attachment
of a chloride anion followed by the loss of HCl, was also
observed. These results indicate the presence of diolate 2 in
solution.
Heating a solution of 2 at 140 °C for 30 min produced
7a,10
Experiments with Methyldioxorhenium(V) MDO. So
far all indications point to reduction of MTO by the sacrificial
alcohol. Therefore, MDO is a reasonable starting complex/
catalyst to investigate. As MDO is not stable without ligation to
ligands other than solvent molecules, we synthesized
MDO·(PPh ) (Figure 2a) and MDO·(PCy ) (Figure 2c)
from the reaction of MTO with PPh and PCy , respectively, in
organic solvent (see Experimental Methods section for details
and single-crystal X-ray structures). MDO·(PR ) complexes
were used as catalysts for the reaction of hydrobenzoin with 3-
octanol at 140 °C. Kinetic measurements were collected
according to method A; the catalyst was added last. As evident
in Figure 2b and d, the reactions proceeded without an
induction period and at a rate comparable to that observed for
reactions starting with MTO. Furthermore, the dependence on
MTO, MDO·diolate,
and benzaldehyde but not stilbene
(Figure 3b). Observation of ions corresponding to attachment
of chloride anion to MTO, MDO, and MDO·diolate in the
mass spectrometric analysis of the heated solution confirmed
the presence of all three molecules (see Experimental Methods
section and Figure 9). We offer reactions (A) and (B) in
Scheme 2 as plausible pathways accounting for the observed
products. Reaction (A) is triggered by heat, as well as by light at
room temperature. Hence, the novel C−C cleaving reaction
results from heat or light. MTO is formed from intermetallic
3
2
3 2
3
3
3
2
19
oxygen atom transfer, which was confirmed independently by
adding MDO·(PR ) to 2 to give MTO and MDO·diolate
3
2
(Figure 3d and Supporting Information). An important point
herein that should not be overlooked is that despite the
formation and observation of MDO·diolate, trans-stilbene was
not produced. Therefore, it can be ascertained that
MDO·diolate in the absence of a reductant does not extrude
alkene by itself. Nevertheless, if the mixture of 2 and
MDO·(PR ) is heated for a prolonged time (30 min), trans-
[
hydrobenzoin] was zeroth-order. These findings bode well
with the hypothesis that MDO is involved in the catalytic cycle,
but once it is formed, it does not return to MTO. In other
words, MDO can be on the catalytic path but not via a cycle
that involves MTO as suggested in Scheme 1.
Oxorhenium(VII) Diolate (2) and Its Potential Involve-
ment. 2 was synthesized by the reaction of MTO and (R,R)-
3
2
stilbene is eventually produced but only after all of the
rhenium(VII) diolate (2) has been consumed (Figure 4). Such
an observation is consistent with autocatalysis. Furthermore,
the rate constant for the formation of trans-stilbene from this
(
+)-hydrobenzoin (1:1) in CDCl in the presence of molecular
3
17
−1
sieves at room temperature. The rhenium(VII) diolate was
obtained in higher purity if the reaction was carried out in the
dark by covering the flask with aluminum foil. The rhenium-
mixture is quite slow (<0.000 58 s ) compared to the turnover
−
1
frequency of a typical catalytic reaction, 0.0278 s . In
conclusion, Scheme 1, which involves the formation of 2
from MTO and the extrusion of stilbene from MDO·diolate, is
refuted.
(
VII) diolate (2) undergoes photodecomposition at room
temperature (Scheme 2, reaction A). Complex 2 was
When 3-octanol was added to the mixture containing MTO
and MDO·diolate, at 140 °C, trans-stilbene was produced along
with 3-octanone (Figure 3c) within 30 min. On the basis of
these observations, we propose that the MDO·diolate must be
further reduced to form a rhenium(III) diolate, and the latter is
responsible for alkene extrusion and regeneration of MDO
Scheme 2. Reactions of Rhenium Diolate at 140 °C
(
reaction (C) in Scheme 2).
The driving force is sufficient with alcohol under our
conditions (140 °C) to form rhenium(III) diolate. Ison and co-
workers have shown recently the formation of rhenium(III)
acetate by reduction of oxorhenium(V) acyl with CO under
20
mild conditions. Interestingly, a recent computational study
of the DODH reaction with MTO using H as well as alcohols
2
as reductants did not consider the involvement of rhenium-
21
(
III).
MDO is produced from the reaction of MTO and 3-octanol,
accounting for the induction period observed when MTO is
added last. Espenson et al. have shown that MDO could form a
22
dimer D1. Our own findings show further reduction to
rhenium NPs at prolonged incubation at high temperature in 3-
octanol. Carbon−carbon bond cleavage in rhenium(VII)
diolate can also lead to MDO formation. MDO·diolate (4)
reacts further with 3-octanol to provide a transient rhenium-
(III) that extrudes alkene and regenerates MDO. In addition,
Re(V) diolate could also react with Re(III) diolate to form the
1
characterized by H NMR. It shows resonances at 5.7 ppm
for the diolato ring protons and a singlet at 2.7 ppm for the Re-
CH (Figure 3a). The structure of 2 was also characterized by
mass spectrometric analysis of the reaction mixture in CDCl , a
technique that has been used previously to study catalytic
intermediates. Upon ionization by atmospheric pressure
chemical ionization (APCI), ions corresponding to attachment
of a chloride anion to the diolate 2 were observed (see
3
3
1
8
dinuclear complex D or with itself to form a rhenium(V)
2
diolate dimer D . The evidence for these dinuclear rhenium
3
E
dx.doi.org/10.1021/om400127z | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
1
1
1
Figure 3. (a) H NMR spectrum of Re(VII) diolate (2); (b) H NMR spectrum of a solution of 2 after heating at 140 °C for 30 min; (c) H NMR
1
spectrum of the mixture in (b) with 3-octanol at 140 °C for 30 min; (d) H NMR spectrum of the OAT reaction between Re(VII)·diolate (2) and
MDO·(PR ) to give MDO·diolate·PR and MTO. (The spectrum shown was acquired during the course of the reaction before completion. For
3
2
3
more details, please refer to the Supporting Information.)
discussed in detail in a later section. The improved mechanism
is shown in Scheme 3.
Rate-Determining Step. A series of NMR experiments
were carried out at high catalyst loading (10 mol % MTO) to
1
characterize catalytic species. The H spectra at different times
(
2, 4, 6, 8, and 10 min) during the course of the reaction are
displayed in Figure 5a. In addition to the methyl peak of MTO
CH ReO ) at 2.5 ppm two other singlets are observed at 3.07
(
3
3
and 2.66 ppm, which are assigned to MDO·diolate (4), or one
VII
of its dinuclear forms D2 and D , and Re ·diolate (2),
3
respectively. Since the peak at 3.07 ppm is larger than the other
two for the majority of the reaction, it is reasonable to assume
that the reduction of MDO·diolate is the rate-determining step.
Buildup in MTO as the reaction progresses is observed in open
air and is attributed to oxidation of MDO and/or MDO·diolate
under the reaction conditions (140 °C). For example, when the
reaction is run with MDO·(PCy ) as the starting catalyst under
3
2
inert atmosphere (N or Ar), only 3.07 ppm (MDO·diolate or
Figure 4. Kinetic profile for a heated solution of Re(VII)·diolate 2 at
40 °C over the course of ca. 2 h. The open circles represent the
2
its dinuclear forms D and D ) is observed along with ca. 37%
1
2
3
disappearance of Re(VII)·diolate 2, which follows first-order depend-
ence in [2] (solid line). The hatched squares represent trans-stilbene
formation.
Re(VII)·diolate (2) (Figure 5b). Re(VII)·diolate is generated
from slow disproportionation of MDO·diolate or its dinuclear
compounds.
To discern the involvement of 3-octanol in the rate-
determining step, we determined the kinetic isotope effect
(KIE) using 3-D-octanol. The results are listed in Table 2. A
complexes emerges from the very unique and unusual rate
dependence on total catalyst concentration [Re]T and is
F
dx.doi.org/10.1021/om400127z | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Figure 7. Thermal ellipsoid plots at the 50% level (H atoms have been
omitted for clarity, in addition the PPH and PCy ligands have been
3
3
depicted in wireframe) of (a) MDO·(PPh ) . Selected bond lengths
3
2
(
Å) and angles (deg): Re1−O1, 1.745(2); Re1−O2, 1.780(3); Re1−
C1, 2.114(5); Re1−P1, 2.4631(7); Re1−P2, 2.4691(6); P1−Re1−P2,
73.06(2); P1−Re1−O1, 86.35(7); P1−Re1−C1, 91.97(13); C1−
Re1−O1, 114.68(18); O1−Re1−O2, 139.19(16). (b) MDO·(PCy ) .
1
3
2
Selected bond lengths (Å) and angles (deg): Re1−O1, 1.7493(10);
Re1−O2, 1.7491(10); Re1−C1, 2.1294(15); Re1−P1, 2.4797(3);
Re1−P2, 2.4881(3); P1−Re1−P2, 163.109(10); P1−Re1−O1,
8
8.69(3); P1−Re1−C1, 98.62(4); C1−Re1−O1, 108.02(7); O1−
Re1−O2, 145.38(5).
1
Figure 5. (a) H NMR spectra acquired every 2 min using 10 mol %
MTO in open air (preheating (R,R)-(+)-hydrobenzoin with 10 equiv
3
-octanol to 140 °C, followed by addition of MTO (method A)); (b)
1
H NMR spectra of the reaction starting with MDO(PCy ) under Ar
3
2
(
1.25 mmol of (R,R)-(+)-hydrobenzoin with 12.5 mmol of 3-octanol
and 10 mol % MDO(PCy ) ). Spectrum shown was acquired after 5
3
2
min.
Figure 8. Mass spectrum of Re(VII) diolate 2 (formation of ReO − is
4
probably due to oxidation occurring under the APCI conditions).
Figure 9. Mass spectrum of the mixture formed upon heating a
Re(VII) diolate (2) solution at 140 °C for 30 min, showing the
formation of MDO, MDO-diolate, and MTO.
Figure 6. Reaction rate order dependence on [Re] , using methods A
T
(
open circles) and B (hatched squares).
G
dx.doi.org/10.1021/om400127z | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Scheme 3. Improved Mechanism of the Major Pathways for MTO-Catalyzed DODH Reaction of Vicinal Diol in the Presence of
Sacrificial Alcohol
Table 2. Kinetic Isotope Reactions with Deuterated Alcohol,
with 2 mol % MTO
MTO catalyst last, and method B to addition of hydrobenzoin
last after incubation of MTO in 3-octanol at 140 °C for 10 min.
The rate values were obtained from the slopes of the kinetic
profiles as the reaction follows zeroth-order in [hydrobenzoin].
a
rate for 3-octanol (mol L−1 s−1
)
rate for D-3-octanol (mol L−1 s−1
)
KIE
1.4
.56 × 10−
4
3.26 × 10−4
4
Plots of log of the rate versus log [Re] revealed 0.49 and 0.65
T
a
Both rates are measured as the slopes of reaction profiles according to
order dependence in catalyst (Figure 6). This observation is in
method A, with MTO added last.
agreement with our hypothesis that MDO·diolate (4) forms a
dinuclear complex with either Re(III) diolate to give D or itself
2
KIE of 1.4 is consistent with the reduction of MDO·diolate
being rate-determining. The value of KIE being <2 can be
rationalized in terms of alcohol coordination followed by H/D-
transfer. A measure of KIE for the stoichiometric reaction of
to give D3 (Scheme 3). Either scenario is kinetically
indistinguishable and manifests itself by giving a half-order
dependence on [Re] . We illustrate herein the rate law
T
derivation for involvement of D and using the rate constants
2
MDO(diolate)(PCy ) also yielded a KIE of 1.2, consistent
given in Scheme 3 for the main catalytic cycle: R = −d[diol]/
3
1
within experimental error with that observed for the catalytic
reaction.
Thermodynamic activation parameters were determined
from kinetic measurements at different temperatures, from
dt = k [MDO][diol]; R = k [MDO diolate], alcohol is the
1
2
2
solvent and thus aalcohol = 1; R = k [Re(III) diolate].
3
3
Assuming that all steps in the catalytic cycle contribute and
R ≈ R ≈ R ,
1
2
3
4
05 to 428 K (132−155 °C). The rate values obtained are as
−1
−1
−1 −1
Re(III)diolate] = kk 23 [MDO diolate]
follows: 0.000 32 mol L
s
at 405 K, 0.000 46 mol L
s
at
[
−1
−1
−1 −1
413 K, 0.000 72 mol L
s
at 421 K, and 0.000 93 mol L
s
at 428 K. Activation parameters were obtained by least-squares
‡
fittings of the data to the Eyring equation, resulting in ΔH =
k4
5 ± 14 kJ mol−1 and ΔS = −154 ± 33 J mol K . The large
‡
−1 −1
[D2] =
[MDO diolate][Re(III) diolate]
6
k
−
4
and negative entropy of activation is in good agreement with a
bimolecular transition state.
k k
4
2
[MDO diolate]2
=
Reaction Rate Order in Rhenium/Catalyst. Kinetic
measurements of the reaction of hydrobenzoin in 3-octanol
at different concentrations of MTO were carried out to
determine the dependence on catalyst concentration. The
results are listed in Table 3. Method A refers to addition of the
k−4k3
The mass balance is
[
Re(III) diolate] + [MDO diolate] + [D ] + [MDO]
2
=
[Re]T
Table 3. Rate Constants with Different [MTO] Using
Methods A and B
[
[
MDO] ≈ 0, and if we assume that [Re(III) diolate] and
MDO diolate] are negligible compared to [D ], then
2
[
MTO]
method A rate
method B rate
−1 −1
k k
4 2
(
mol %)
(mol L−
1
s
−1
)
(mol L
s
)
[Re] ≅ [D ] =
[MDO diolate]2
T
2
3.5 × 10−
4
−4
−4
−3
−3
k k
−4 3
1
2
5
4.9 × 10
7.2 × 10
1.6 × 10
2.0 × 10
−
4
4
3
4.6 × 10
k k
−4 3
k k
4 2
6.8 × 10−
1.1 × 10−
[MDO diolate] =
[Re]T
10
H
dx.doi.org/10.1021/om400127z | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Hence, the approximate rate law is
d[diol]
Article
General Procedure for Kinetics. The glycol (2.5 mmol, referred
to (R,R)-(+)-hydrobenzoin in most cases) was mixed with the
sacrificial alcohol (25 mmol, referred to 3-octanol most of the time)
and diphenylmethane (∼0.25 mmol, internal standard) in a three-neck
round-bottom flask. After refluxing the reaction mixture at 140 °C in a
preheated silicone oil bath, CH ReO (0.05 mmol, 2 mol %) was
added, which marked the starting point of the kinetic data collection.
Aliquots were taken at different time points, quenched immediately in
CDCl over ice, and analyzed by H NMR spectroscopy (or GC) to
k k
4 3
−
−
= R = k [MDO diolate] =
[Re]T
2
2
dt
k k
4 2
3
3
This derivation is consistent with the kinetic data as well as
NMR observations that the majority of the catalyst is in a dimer
form (D , D , or D in Scheme 3) and the rate constants k and
1
1
2
3
2
3
k3 contribute to the catalytic rate of the reaction.
generate kinetic profiles of concentration versus time.
V
Synthesis of MeRe O (PPh ) . A solution of MTO (100 mg, 0.40
2
3 2
CONCLUSION
mmol) in 2 mL of dry diethyl ether was treated with a solution of PPh
631 mg, 2.4 mmol) in 3 mL of dry diethyl ether under a nitrogen
3
■
(
We have reported herein a thorough kinetics and mechanistic
study on the deoxydehydration reaction of glycols using 3-
octanol as a reductant and methyltrioxorhenium(VII) as
catalyst. The reaction proceeds smoothly at 140−170 °C
under ambient pressure and open to air to give trans-stilbene
from (R,R)-(+)-hydrobenzoin and 3-octanone as the major
products. The kinetics are zeroth-order in [hydrobenzoin] and
half-order in [Re]. An induction period is observed when the
catalyst is added last. However, if MTO and 3-octanol are
allowed to react for 10 min at the reaction temperature, the
induction period disappears. Two prevalent literature mecha-
nisms were considered, probed, and refuted. Reactions of
rhenium(VII) diolate revealed a novel C−C bond cleaving
reaction that affords methyldioxorhenium(V) and benzalde-
hyde. Oxygen atom transfer from the rhenium(VII) diolate to
MDO yields MTO and MDO·diolate. The latter did not
produce the alkene product unless a reductant was added. All
these findings and additional studies with MDO·(PR ) support
atmosphere. Over 24 h at room temperature, the solution changed
from colorless to deep orange, and the orange product crystallized out
of the diethyl ether solution. The solid was filtered and washed with
dry diethyl ether three times to yield pure MDO·(PPh ) (175.7 mg,
3
2
58% yield). Anal. Calcd for C H O P Re: C, 58.64; H, 4.39. Found:
37
33
2 2
C, 57.66; H, 4.39. The complex is not solution stable. Therefore,
generation of the complex in situ allowed for spectroscopic
characterization (see below).
V
Preparation of MeRe O (PPh ) in Situ. A solution of MTO (31
2
3 2
mg, 0.125 mmol) in 1 mL of CDCl was treated with PPh (98 mg,
3
3
0
.375 mmol). After 1 h 30 min at ∼35 °C or 38 h at room temperature
(
under a nitrogen atmosphere), the solution changed from colorless to
31
yellow to deep red. P NMR analyses showed formation of OPPh3.
31
P NMR (400 MHz,CDCl ): δ 30 (OPPh , 32%), 10 (coordinated
3
3
1
PPh , 41%), −5 (PPh , 27%). H NMR (400 MHz, CDCl ): δ 2.85 (s,
3 H). C NMR (400 MHz, CDCl ): δ 143 ppm.
3
3
3
13
3
V
Synthesis of MeRe O (PCy ) . A solution of MTO (100 mg,
2 3 2
∼0.4 mmol) in 2 mL of dry diethyl ether was treated with a solution of
PCy (675 mg, 2.4 mmol) in 3 mL of dry diethyl ether under a
3
3
2
nitrogen atmosphere. Over 24 h at room temperature, the solution
changed from colorless to deep orange, and the orange product
precipitated out of the diethyl ether solution. The solid was filtered
and washed with dry diethyl ether three times to yield pure
a catalytic cycle in which MDO·diolate is reduced by 3-octanol
to a transient rhenium(III) diolate, which is responsible for
alkene extrusion and regeneration of MDO (Schemes 2 and 3).
The [Re]1 dependence is rationalized by formation of a
/2
V
31
MeRe O (PCy ) (140.3 mg, 44% yield). P NMR (400 MHz,
2
3 2
dinuclear rhenium complex (D ) (Scheme 3). Our investigation
1
2
CDCl ): δ 8 (coordinated PCy ). H NMR (400 MHz, CDCl ): δ 3.0
3
3
3
13
offers a new mechanism for DODH using oxorhenium catalysis
that is quite different from the prevalent rhenium(VII) and
rhenium(V) cycles that have been claimed in the literature
Scheme 1). Our new mechanism should stimulate future
catalyst design on the basis of a rhenium(V)/rhenium(III)
cycle.
(s, 3 H) ppm, 1.1−2.3 ppm (cyclohexyl peaks). C NMR (400 MHz,
CDCl ): δ 140 ppm (coupled to CH ), 34, 29, 27, 26 ppm (cyclohexyl
3
3
peaks). Anal. Calcd for C37
H
69
O
P
2
Re: C, 55.96; H, 8.76. Found: C,
2
5
5.69; H, 8.80.
X-ray Structures of MeRe(O) (PPh ) and MeRe(O) (PCy ) .
(
2
3 2
2
3 2
Details and general procedure for X-ray determination are provided in
the Supporting Information. ORTEP diagrams of both structures with
selected bond lengths and angles are presented in Figure 7.
EXPERIMENTAL METHODS
General Information. All commercial materials were used as
received without further purification unless specified. (R,R)-
■
Mass Spectrometry. All mass spectrometric experiments were
carried out in a Thermo Scientific linear quadrupole ion trap (LQIT)−
Fourier transform ion cyclotron resonance (FT-ICR; 7 T magnet)
mass spectrometer. The nominal pressures of the LQIT vacuum
manifold and the FT-ICR vacuum manifold were maintained at about
(
+)-Hydrobenzoin, 3-octanol, (S,S)-hydrobenzoin, meso-hydroben-
zoin, cis-stilbene oxide, styrene oxide, 1-phenyl-1,2-ethanediol,
erythritol, 2-nonanol, 5-nonanol, 1-heptanol, cyclohexanol, cis-1,2-
cyclooctanediol, trans-1,2-cyclooctanediol, and 1,2-decanediol were
purchased from Sigma-Aldrich. Methyltrioxorhenium, ReO , Re O ,
−5
−10
0.60 × 10 Torr and 2.0 × 10 Torr, respectively, as read by ion
gauges. The LQIT-FT-ICR mass spectrometer was operated using the
LTQ Tune Plus interface.
3
2
7
and NH ReO were purchased from Strem Chemicals. MTO was
An atmospheric pressure chemical ionization (APCI) source was
4
4
1
13
purified by vacuum sublimation at 40 °C. H NMR and C NMR
used for ion generation. Reaction mixtures in CDCl were diluted 100-
3
2
were recorded with a Bruker Avance ARX-400. HMQC and H NMR
fold with dry CDCl in a glovebox. CDCl was used, as it was the
3
3
were recorded with Bruker Avance DRX-500 spectrometers. The GC
analysis was performed using an Agilent Technologies 6890 N with a
solvent for NMR measurements, and it encouraged negative ion
production. The diluted solutions were directly introduced into the
APCI source by using a Hamilton gastight syringe operating at a flow
rate of 20 μL/min. APCI discharge current was 4.0 μA. The vaporizer
temperature was set at 80 or 120 °C for optimized ion signals. Sheath
DB-5 column.
VII
Generation of MeRe O (OCHPhCHPhO) (2). (R,R)-(+)-Hydro-
2
benzoin (25.7 mg, 0.12 mmol) was added to a solution of
methyltrioxorhenium (27.4 mg, 0.11 mmol) in 2 mL of dry methylene
chloride in the presence of molecular sieves. After 24 h at room
temperature (under a nitrogen atmosphere), the solution inside the
NMR tube changed from colorless to yellow to deep red. NMR
and auxiliary gas (N ) flows were 35 and 5 (arbitrary units),
2
respectively. A heated ion transfer capillary/mass spectrometer inlet
temperature of 275 °C was used. All dc voltages and offsets for the ion
optics were optimized utilizing the integrated tuning features of the
LTQ Tune Plus interface. Xcalibur 2.0 software was used for data
analysis. All LQIT mass spectra acquired were an average of 40 scans.
Figure 8 shows MS characterization of MTO-diolate (2), and Figure 9
the reaction mixture resulting from heating complex 2 at 140 °C.
1
analyses showed formation of Re(VII) diolate (2). H NMR (400
MHz, CDCl ): δ 7.10−7.40 ppm (m), 5.75 (br s, 2 H), 2.70 (s, 3 H).
3
1
3
C NMR (500 MHz, CDCl ): δ 128.97, 128.80, 126.54, 108.82, 40
3
ppm (coupled to CH3).
I
dx.doi.org/10.1021/om400127z | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
(
7) (a) Ahmad, I.; Chapman, G.; Nicholas, K. M. Organometallics
ASSOCIATED CONTENT
■
2
011, 30, 2810. (b) Vkuturi, S.; Chapman, G.; Ahmad, I.; Nicholas, K.
*
S
Supporting Information
M. Inorg. Chem. 2010, 49, 4744.
Kinetics details and NMR spectra for the DODH reactions;
UV−vis and TEM images of Re NPs formed in 3-octanol;
synthesis and full characterizations of MDO(PR ) complexes;
(8) Sousa, S. C. A.; Fernandes, A. C. Tetrahedron Lett. 2011, 52,
6960.
(9) Yi, J.; Liu, S.; Abu-Omar, M. M. ChemSusChem 2012, 5, 1401.
3
2
(
10) Shiramizu, M.; Toste, F. D. Angew. Chem., Int. Ed. 2012, 51,
082.
11) (a) Thibault, M. E.; DiMondo, D. V.; Jennings, M.; Abdelnur, P.
X-ray characterizations of MDO(PR ) (R = Ph and Cy)
3
2
8
(
CIF); preparation and full characterization by NMR including
(
HMQC of MTO(diolate) and MDO(diolate)(PCy ) ; details
3
2
V.; Eberlin, M. N.; Schlaf, M. Green Chem. 2011, 13, 357. (b) Schlaf,
M.; Ghosh, P.; Fagan, P. J.; Hauptman, E.; Bullock, R. M. Adv. Synth.
Catal. 2009, 351, 789. (c) Ghosh, P.; Fagan, P. J.; Marshall, W. J.;
Hauptman, E.; Bullock, R. M. Inorg. Chem. 2009, 48, 6490.
and spectral characterization on the inert-metal OAT between
MTO(diolate) and MDO; and results on DODH in the
absence of a sacrificial alcohol. This material is available free of
charge via the Internet at http://pubs.acs.org.
(
12) Stanowski, S.; Nicholas, K. M.; Srivastava, R. S. Organometallics
012, 31, 515.
13) (a) Zhu, Z. L.; Espenson, J. H. J. Org. Chem. 1996, 61, 324.
b) Herrmann, W. A.; Roesky, P. W.; Wang, M.; Scherer, W.
2
(
(
AUTHOR INFORMATION
■
Corresponding Author
Organometallics 1994, 13, 4531. (c) Gable, K. P.; Phan, T. N. J. Am.
Chem. Soc. 1994, 116, 833. (d) Gable, K. P.; Juliette, J. J. J. J. Am.
Chem. Soc. 1995, 117, 955. (e) Gable, K. P.; AbuBaker, A.; Zientara,
K.; Wainwright, A. M. Organometallics 1999, 18, 173.
*E-mail: mabuomar@purdue.edu.
Present Address
§
Department of Chemistry, College of Science, Gaziosmanpasa
(
14) Korstanje, T. J.; Jastrzebski, J. T. B. H.; Gebbink, R. J. M. K.
ChemSusChem 2010, 3, 695.
15) (a) Biswas, K.; Rao, C. N. R. J. Phys. Chem. B 2005, 110, 842.
b) Ghosh, S.; Biswas, K.; Rao, C. N. R. J. Mater. Chem. 2007, 17,
412.
16) (a) Gomez, M.; Favier, I. “Metal Nanoparticles Dispersed in
University, Tasliciftlik, 60240, Tokat, Turkey.
(
(
2
Author Contributions
All authors have given approval to the final version of the
manuscript.
(
Funding
Solution: Tests to Identify the Catalyst Nature. In Metal Nanoclusters
in Catalysis and Materials Science: The Issue of Size Control; Corain, B.;
Schmid, G.; Toshima, N., Eds.; Elsevier, 2008; p 427. (b) Borowski, A.
F.; Sabo-Etienne, S.; Chaudret, B. J. Mol. Catal. A: Chem. 2001, 174,
Notes
The authors declare no competing financial interest.
6
9. (c) Georgiades, G. C.; Sermon, P. A. J. Chem. Soc., Chem. Commun.
1985, 975.
17) (a) Zhu, Z. L.; AlAjlouni, A. M.; Espenson, J. H. Inorg. Chem.
996, 35, 1408. (b) Haider, J. J.; Kratzer, R. M.; Herrmann, W. A.;
Zhao, J.; Kuhn, F. E. J. Organomet. Chem. 2004, 689, 3735.
18) (a) Zhang, X.; Chen, X.; Chen, P. Organometallics 2004, 23,
3437. (b) Jackson, S. M.; Chisholm, D. M.; McIndoe, J. S.; Rosenberg,
L. Eur. J. Inorg. Chem. 2011, 2011, 327.
(19) (a) Conry, R. R.; Mayer, J. M. Inorg. Chem. 1990, 29, 4862.
ACKNOWLEDGMENTS
■
(
1
S.L. and M.M.A.-O. wish to thank Dr. Grigori Medvedev and
Ms. Silei Xiong of Purdue University for useful discussions on
kinetics. A.S. acknowledges support from TUBITAK Interna-
tional Postdoctoral Research Scholarship Programme. E.A.I.
acknowledges the National Science Foundation via the
CAREER Award (CHE-0955636) for funding. L.Y. and the
described MS studies were supported as part of the Center for
Direct Catalytic Conversion of Biomass to Biofuels (C3Bio), an
Energy Frontier Research Center (EFRC) funded by the U.S.
Department of Energy, Office of Science, Office of Basic Energy
Sciences, under award number DE-SC-0000997. This research
was supported by the U.S. Department of Energy, Basic Energy
Sciences (DOE-BES grant number DE-FG-02-06ER15794).
̈
(
(b) McPherson, L. D.; Drees, M.; Khan, S. I.; Strassner, T.; Abu-Omar,
M. M. Inorg. Chem. 2004, 43, 4036. (c) Ison, E. A.; Cessarich, J. E.;
Travia, N. E.; Fanwick, P. E.; Abu-Omar, M. M. J. Am. Chem. Soc.
2
(
1
2
(
3
(
007, 129, 1167.
20) (a) Smeltz, J. L.; Boyle, P. D.; Ison, E. A. J. Am. Chem. Soc. 2011,
33, 13288. (b) Lilly, C. P.; Boyle, P. D.; Ison, E. A. Organometallics
012, 31, 4295.
21) Qu, S.; Dang, Y.; Wen, M.; Wang, Z.-X. Chem.−Eur. J. 2013, 19,
827.
22) (a) Abu-Omar, M. M.; Appelman, E. H.; Espenson, J. H. Inorg.
REFERENCES
■
Chem. 1996, 35, 7751. (b) Herrmann, W. A.; Fischer, R. W.; Scherer,
W. Adv. Mater. 1992, 4, 653. (c) Hermann, W. A.; Fischer, R. W. J. Am.
Chem. Soc. 1995, 117, 3223. (d) Espenson, J. H.; Yiu, D. T. Y. Inorg.
Chem. 2000, 39, 4113.
(
1) (a) Huber, G. W.; Iborra, S.; Corma, A. Chem. Rev. 2006, 106,
044. (b) Serrano-Ruiz, J. C.; Luque, R.; Sepulveda-Escribano, A.
4
Chem. Soc. Rev. 2011, 40, 5266. (c) Alonso, D. M.; Bond, J. Q.;
Dumesic, J. A. Green Chem. 2010, 12, 1493. (d) Gallezot, P. Chem. Soc.
Rev. 2012, 41, 1538. (e) Huber, G. W.; Corma, A. Angew. Chem., Int.
Ed. 2007, 46, 7184. (f) Chheda, J. N.; Huber, G. W.; Dumesic, J. A.
Angew. Chem., Int. Ed. 2007, 46, 7164.
(
2) (a) Herrmann, W. A.; Ku
b) Rudolph, J.; Reddy, K. L.; Chiang, J. P.; Sharpless, K. B. J. Am.
Chem. Soc. 1997, 119, 6189. (c) Jonsson, S. Y.; Farnegardh, K.;
Backvall, J.-E. J. Am. Chem. Soc. 2001, 123, 1365.
3) (a) Gable, K. P. Organometallics 1994, 13, 2486. (b) Zhu, Z.;
̈
hn, F. E. Acc. Chem. Res. 1997, 30, 169.
(
̈
̊
̈
(
Espenson, J. H. J. Mol. Catal. A: Chem. 1995, 103, 87. (c) Gable, K. P.;
Brown, E. C. Organometallics 2000, 19, 944.
(
(
4) Cook, G. K.; Andrews, M. A. J. Am. Chem. Soc. 1996, 118, 9448.
5) Ziegler, J. Z.; Zdilla, M. J.; Evans, A. J.; Abu-Omar, M. M. Inorg.
Chem. 2009, 48, 9998.
6) Arceo, E.; Ellman, J. A.; Bergman, R. G. J. Am. Chem. Soc. 2010,
32, 11408.
(
1
J
dx.doi.org/10.1021/om400127z | Organometallics XXXX, XXX, XXX−XXX